feat(datasets) Add pathological partitioner #3623
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
adam-narozniak
requested review from
jafermarq,
tanertopal and
danieljanes
as code owners
jafermarq
reviewed
Jul 10, 2024
jafermarq
reviewed
Jul 10, 2024
datasets/flwr_datasets/partitioner/pathological_partitioner_test.py
Outdated
Show resolved
Hide resolved
jafermarq
reviewed
Jul 10, 2024
adam-narozniak
changed the title
jafermarq
approved these changes
Jul 12, 2024
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Issue
There's no support for what is commonly known as a
Pathologicalpartitioner.Description
PathologicalPartitioner(the version that does not take the distribution as an input) provides a way of creating partition IDs such that the number of unique classes is constrained to the number specified by users. Note that it does not imply that each class is divided evenly (see more in docs).This method is used in our baselines in:
Proposal
Provide an implementation of
PathologicalPartitioner. For details on how the method works, please refer to the documentation and ask questions directly there.Explanation
Note: if reproducing these images, they will look slightly different since a change that sorts the unique labels was applied after the images were generated.
Here's the visualization of the results that can be obtained using
PathologicalPartitioner(not available in the docstrings but can help to understand the method):